The Fall of the Dollar
First, there was rapid growth in exports across Asia with a concurrent rise in their trade surplus. These goods and services were artificially cheap to US consumers. This export surge also occurred in major commodity producing countries as they revved up to support the export-driven industrial boom occurring. Some commodity producers directly pegged themselves to the dollar, like the Middle East, resulting in further recycling flows.
The second effect, caused by the first, was a tremendous accumulation of dollar reserves by pegging countries. These dollars were invested by those countries into risk-free assets, driving down real interest rates on longer Treasuries and leading to Greenspan's famous conundrum. Private investors moved to spread assets and bid them far beyond rational prices, compressing spreads and interest rates across the risk spectrum.
These recycling flows were vendor financing on a massive scale. Without the pegs, US real interest rates would have soared, halting the process. Instead, a tremendous drop in real interest rates caused by this vendor financing enabled US consumers and corporations to take out stunning amounts of debt via HELOC's, asset bubbles, cov-lite loans, and so forth.
Through this financial intermediation, more and more money was available to the US to purchase commodities and finished goods from the rest of the world. The rest of the world bought more Treasuries and other safe debt, causing asset values to soar and interest rates to drop. This was all a positive feedback loop.
A lot of people saw the massive deficits being run by the US and thought the USD must inevitably crash. The consensus was that the USD would crash should China cease its massive reinvestment of its surplus. Indeed, the USD weakened gradually for many years in a row.
Compiling this, I posit a completely different explanation. The weak USD was a result of the vendor financing. As real interest rates were abnormally compressed by return-insensitive central banks, private capital fled the US to other destinations in search of higher real returns. Because USD was unable to weaken against CNY, it weakened against other currencies instead, making the real returns available even worse.
The strength of EUR and GBP was partially an export valve for some of the pressure on USD/CNY. As EUR and GBP strengthened against the USD, the asset bubble spread, and they picked up the consumer of last resort role from America's weary citizens.
The Rise of the Dollar
I first became concerned in February that the Earth's magnetic pole had flipped, making this powerful engine turn in reverse.
The weakness of the USD did begin to turn around in March, as the financial crisis began to affect trade in finished goods. It finally truly reversed upwards with the final bursting of the commodity bubble, and it's strengthened ever since.
Everyone is terrified the USD will continue to fall, mainly as a result of the massive quantitative and qualitative easing schemes underway by the Fed, the persistent current account deficit, the U.S. NIIP position, and so forth. I'm scared too, and I want to remain that way. But I can't justify my fear, and haven't been able to for awhile.
The USD cannot weaken against the CNY or JPY, and I'm very skeptical that there is any positive traction whatsoever from the fiscal and monetary programs underway. While there is a great deal of disagreement about what determines exchange rates, it's probable that relative real interest rates matter far more than money supply. And those are only moving in the USD's favor right now.
Indeed, government deficits raise real interest rates, rather than suppressing them. This is intuitively obvious and empirically demonstrable, but theoretically indeterminate, which confuses economists everywhere. Whether that is true in a liquidity trap environment, or whether we're in a liquidity trap now, remains up for debate. But I don't see any reason to believe, outside a stylized AD/AS curve and baseless Keynesian multipliers, that the rules should not still hold.
The effects of an insolvent Fed's debt load are the same as that for Treasury debt, which means that all these stimulus and rescue programs, including direct monetization, should only lead to higher real interest rates. Since the dollar can't lose value right now by imposition of the pegging countries, and nominal interest rates have hit the zero bound, I conclude that fiscal stimulus and debt monetization will strengthen the dollar and worsen deflation. Until the pegs are broken or serious inflation erupts in the rest of the world, this is the only possible outcome.
We have seen a further collapse in trade since September which is only worsening, lessening the odds of these pegs breaking, and reducing further the vendor financing that can be recycled. I expect to see further increases in US real interest rates.
The USD should continue to strengthen until serious inflation occurs in China, Japan, and other countries, or the pegs break. Until such a time, due to the increase in real interest rates and the damage inflicted by deflation, most assets, and particularly longer-dated ones, will decline in value.
If the pegs never break, and inflation never erupts overseas, then we can expect a massive wave of defaults in the US, possibly including Treasury. So, in perhaps the most likely outcome, USD assets could crash while the USD pulls through just fine.
Posts
14 comments:
Well written.
I have a couple of comments on the backdrop to your subject, without addressing your main theme of the dollar:
I’ve always been sceptical of the real interest rate argument. Most link it directly to the quantity of capital inflows. You link it indirectly (?) to the pegs. But without the pegs, the surpluses would have been lower and the natural supply of net capital inflows purchasing financial assets that much less – the counterfactual. Part of my dislike for the real interest rate argument is that it ignores these other variables that are different from the counterfactual. There’s a circularity to the argument. Vendor financing wouldn’t have been necessary at all without the vending. In this sense, the flows alone don’t explain the level of real rates. Something else must.
The more refined explanation is something called “the global savings glut”. I’ve always had a problem with this as well. The problem is that the argument lacks intellectual integrity and honesty. I’ve seen the stage set variously for explaining the glut in terms of current account surpluses (which are not global), the “difference” between global savings and global investment (which is global, but very debatable in theory, and not supported by data and in fact contradicted in terms of data by Stephen Roach), and the fact that its mostly central bank financing rather than private financing. Which is it? These are not intersecting dimensions. The usual suspects/proselytizers hopscotch from one independent rationale to the next in an endless cycle of trying to frame an argument that has no honest reference point.
(I have no real problem with your explanation, which is in the form of a reasonable story.)
Dang, you're fast, JKH.
Most link it directly to the quantity of capital inflows. You link it indirectly (?) to the pegs.
Depending on where the pegs are set relative to equilibrium, these two are one and the same. An exceedingly cheap peg means a lot of capital flow.
Vendor financing wouldn’t have been necessary at all without the vending.
I might not be fully grokking your point, but I simply think these flows were the result of a policy decision. After the trauma of 1997, there was clearly a pan-Asiatic decision to accumulate reserves and create as much exports as possible, to make sure 1997 never happened again. These very low pegs were an obvious and excellent cudgel for the job. That was enough to create both the vending and the vendor financing.
In this sense, the flows alone don’t explain the level of real rates.
Chinn has done interesting empirical work on the magnitude of the suppression, and I linked to it somewhere in that pile of references. IIRC, he found long real interest rates in the US were depressed by recycling flows to the tune of 1%, which is a pretty large effect. I bet that was sufficient fuse for the financial intermediation bomb to cause the very low rates we saw.
I’ve seen the stage set variously for explaining the glut in terms of current account surpluses (which are not global), the “difference” between global savings and global investment (which is global, but very debatable in theory, and not supported by data and in fact contradicted in terms of data by Stephen Roach), and the fact that its mostly central bank financing rather than private financing.
Roach's analysis was the last straw for me too with the global savings glut theory. I never believed it, and now I actually viscerally dislike it.
Yes, studies have “demonstrated” that capital inflows have resulted in low real rates.
Suppose study X concludes that net capital flows into the US of $ 800 billion have resulted in a decline of 1 per cent in the real rate.
The question I have is – compared to what?
Compared to a world in which there is no net capital inflow of $ 800 billion? Well, that’s a world where the US has no current account deficit and no requirement for net capital inflows. It’s a completely different world. Why would real interest rates have been higher in a world in which US financing needs are $ 800 billion less?
Of course, the savings glut theory with great flexibility of rationale turns this question on its head and responds this way:
Well, real rates are low; therefore they must be evidence of a savings glut.
Couple of questions:
1.Yen carry trade?
2. The yen is strengthening against the dollar - with some calling for a 50 handle (down form 160 in 1990)? Once the export regime breaks (Tiawan exports down 40% most recently) the incentive to perpetuate the system goes away. Once this happens the incentive moves to destroy the dollar or change the regime (terms of trade) to a basket / other and hence break the $ commodity monopoly. Seems to me the argument is not abut dollar strength in the near term against yen or yuan. It is instead about what the next regime looks like and how the dollar is positioned in that regime. Kissinger's piece in IHT speaks to this revised world order in a more political hue. He espouses a structure or a return to Bretton Woods like regime as compared to an all out protectionist race to the bottom. What does this mean for the US? For sure US will fight it tooth and nail. But discipline will be imposed on the dollar which can only in the long run add tothe downward pressure - as wage differential will persist indefinitly
3. A controlled demolition in the form of a negotiated new regime which imposes some constraints on the dollar medium term seems most likely. I don't think the detonation is avoidable - but it s in everyone interest to see that a new regime is in place before this one finally ends.
3. The dollar may not lose value against the pegs but it could lose value against "things." Yes deflation is paramount at the moment, but it is a near certainty that as the anglo countries attempt to print their way out they seal the fate of the dollar hegemony (already sealed) on a going forward basis.
4. WIthout the pegs (or a concerted move by the US to balance trade) there would exist no incentive to keep the artificial scheme going. Would not not the logical extension be that once this happens real rates would explode and attract capital flows? Reminds me of the arguments about the yuan revaluation: as people looked to diversity out of China (free capital) the revaluation would not be nearly as severe as some suggest. Circularity.
3. A controlled demolition...
I find this implausible because it makes sense and entails truth telling.
The question I have is – compared to what?
JKH, in a discipline where it's impossible to conduct explicit experiments or run controls, this question will always be present. I'd rather not let it paralyze analysis. Expected behavior based on past performance is all we've got, and better than nothing, as long as viewed with appropriate skepticism.
The yen is strengthening against the dollar - with some calling for a 50 handle (down form 160 in 1990)?
Thanks for your comments, S. This would be a welcome and positive development, although the pace may be insufficient for meaningful real rate adjustment given Japan's endemic deflation problem.
Once the export regime breaks (Tiawan exports down 40% most recently) the incentive to perpetuate the system goes away.
Does it break, or does it get stronger? I'm not sure. Remember that everyone has to agree to revalue their currency stronger, not just China. China is the Saudi Arabia of the manufacturing OPEC, but they still face all the enforcement problems OPEC does.
But discipline will be imposed on the dollar which can only in the long run add tothe downward pressure - as wage differential will persist indefinitly
If the dollar goes down, the wage differential can start to adjust too. This would be a very positive development.
The dollar may not lose value against the pegs but it could lose value against "things."
Remember the impact of real interest rates on commodities. I think the USD is more likely to gain value against things, leading to structural underinvestment in things at some point in the future, at which point things can begin to recover.
I recognize that this is a way-out-of-consensus call.
Would not not the logical extension be that once this happens real rates would explode and attract capital flows?
I think this is a very powerful argument, S, but I have no way of really knowing what the outcome of a free float would be. Certainly we've seen from the hot money reversal that the yuan does not always and everywhere rise, and China's currency reserves aren't ever-growing. But they still have a large structural current account surplus, so it's probably somewhat stronger than today's levels. Very interesting to ponder how much.
“In a discipline where it's impossible to conduct explicit experiments or run controls, this question will always be present. I'd rather not let it paralyze analysis. Expected behaviour based on past performance is all we've got, and better than nothing, as long as viewed with appropriate scepticism.”
I really don’t like to harp on this because it distracts from your main theme.
But one last try (I promise). I like your answer in terms of the question of pursuing the investigation of the counterfactual. But that’s not my question. The question involves the consideration of the counterfactual in a way that is consistent with the conclusion regarding actual events. In this sense, the question becomes the following. Why is it the case that the real rate discount that is presumed to be the result of capital inflows (and therefore also the result of a current account deficit) would not be present in a world without that particular imbalance? In other words, how do you know that the real rate effect is the result of this particular imbalance as opposed to other factors? I’m not asking for measurement. I’m asking for proof in the form of denial of plausibility. I’ve not seen anybody answer this question. Otherwise it’s just a theory of convenience.
Appreciate your patience on this point. I’m only asking because I think you may have as good a shot at the answer as anybody I’ve read.
Appreciate your patience on this point. I’m only asking because I think you may have as good a shot at the answer as anybody I’ve read.
I've got a good idea what you're asking now. Let me try rephrasing it succinctly:
"What was the equilibrium real interest rate over the period of the last decade, and if the real interest rate deviated dramatically from equilibrium, can we exclude major factors other than vendor financing flows from consideration?"
If I've got it right, it's an extremely relevant and difficult question, JKH. I don't think I'm particularly poised nor qualified to answer it, but I'll take a look at it over the weekend, since I've got a lot of real job stuff to do through the balance of this week. If I can arrive at an answer that I'm comfortable with, I'll do a post. No guarantees.
For now, all I can say is that the mechanism I propose correlates well with the data, has strong historical empirical support, and has good theoretical underpinnings.
Wish me luck, and bring an easier one with you next time...
@NDK,
Your scenario appears highly plausible to me and I agree with most of what you have laid out. However, I take slight issue with that the stimulus and rescue programs must lead to higher real interest rates. I would certainly agree with this from a new classical standpoint.
However, this kind of a crowd out scenario entails the existence of plausible (to investors) alternatives in the private markets. I am taking a bit of a jump here, but current junk/AAA/gov bond spreads imply there is a substantial premium given to government bonds based upon investor risk expectations/assessments. The investor is willing to pay a large premium, rational or otherwise, to hold a "riskless" security. While I feel that US Treasurys yields will need to grow to attract more capital for a variety of reasons, we have not seen evidence of this yet. The recent treasury auctions have been going well considering the inflationary rhetoric the FED is talking and that is implied by their actions (and of the US government). Perhaps this investor sentiment remains throughout the entirity of the coming stimulus packages, perhaps not. All I am trying to assert is that treasury yield appreciation is not necessarily bound to occur. But the case you have laid out for such an occurence is quite strong.
Another factor to examine will be the ability of the vendor countries to physically send the US enough goods to offset our currency depreciation efforts. I strongly feel that there is not enough shipping capacity to be able to offset the inflationary measures we are incurring. Perhaps they continue to buy dollar assets to allow us to take out more debt and nullify our keynesian multiplier (if they exist in the first place), but the amount of ships going to breakers coupled with the amount going into disrepair and new orders being canceled implies a low inventory of active ships. In this environment in may be very difficult to lay in enough Twn/Chi/Jap goods to halt US dollar devaluation. This does not even mention production capacity in China being permantly scrapped and the friction labor will encounter as it returns from the chinese countryside to the factories. Who wins the race, the printing press or the ship captain?
At the end of the day, what value do these exported dollars have to the vendor nations if we print enough to devalue our currency? We hold artificially low priced goods, they hold worthless pieces of paper. Or, logically, wouldn't the dollars trickle back into our system over time from their purchases? Could these dollars repatriate to the US in the form of investment in capital production equipment or infrastructure investments from abroad? Do we devalue our dollar so much and export so much we attract foriegn USD holders to invest in real US assets?
Perhaps these trade/production frictions will last long enough for debt repayment. It is also possible that the US forces China and Japan to halt their peg. Perhaps we impose protectionist measures to halt the export of our inflation. These tariffs may counteract artificially low currency valuations and are plausible (considering you convince people the current Yuan/Yen situation is NOT FREE TRADE).
Either way, it will be very interesting how this crisis plays out through agent/government action/reaction.
I hope you continue to focus on this subject. Whither the dollar and dollar based assets, is, and will be, the central question of our time.
@ ndk
Do you think that the Fed's apparent commitment to inflation, with the help of the treasury, will help to increase inflation expectations and reduce those high real interest rates? I think that this theory may hold water, as Greg Mankiw and Paul Krugman seem to have come to a consensus on the matter (refer to http://krugman.blogs.nytimes.com/2008/12/17/a-whiff-of-inflationary-grapeshot/).
Thanks in advance for your thoughts!
-Joe Hill
Excellent feedback, RPB.
The investor is willing to pay a large premium, rational or otherwise, to hold a "riskless" security. While I feel that US Treasurys yields will need to grow to attract more capital for a variety of reasons, we have not seen evidence of this yet.
We haven't seen nominal yields rise much, but real yields are a different story entirely. Those have risen substantially for all asset classes. I don't really expect nominal yields to rise much until the pegs break and/or inflation erupts in the pegging countries, for the reasons set out in my original post.
Some of the more aggressive monetization by the Fed has indeed driven yields lower on, for example, FNM/FRE MBS. While I believe we can lower real interest rates in a specific sector of the market, I don't think it's a good idea, first for the misalignments created, and secondly because an equal amount of displacement upwards in real interest rates elsewhere would be expected, again from that neoclassical perspective you mention.
As an aside, I'm skeptical of the pure delineation some assume into existence when they speak of no or limited crowding out in the real economy from stimulus spending. Eugene Fama puts the argument forward more eloquently than I can.
Another factor to examine will be the ability of the vendor countries to physically send the US enough goods to offset our currency depreciation efforts. I strongly feel that there is not enough shipping capacity to be able to offset the inflationary measures we are incurring.
This is an interesting perspective, given the limits on intervention imposed by the trade deficit. There is certainly such a limit, and it's coming down. With an upper limit on the size of the trade deficit that can be run, there is an accompanying upper limit on the amount of USD assets purchased by our trading partners.
But to the extent that prices can be forced upwards through excessive artificial demand, those higher prices paid result in the accrual of more surplus by trading partners, i.e. putting the money directly into the pockets of the peggers. I don't see the mechanism that would translate that rise into a sustainable gain in wages or prices, much less a self-perpetuating move upward.
Such artificial price subsidies would only encourage more investment in tradeables overseas, more investment in ships, and less investment in exposed domestic sectors.
At the end of the day, what value do these exported dollars have to the vendor nations if we print enough to devalue our currency?
I have a very realpolitik answer to this that I've tried not to emphasize. Give a man a fish, and he'll eat it. Give a man fish for years in a row, and he'll stop fishing himself. The stream is yours.
I don't believe it's the full answer, nor do I believe the openly stated benevolent answer, but think the truth lies somewhere in between. I could write about it at length someday, I guess.
Could these dollars repatriate to the US in the form of investment in capital production equipment or infrastructure investments from abroad?
China and Japan have both already floated this idea. China did so in a punitive way, wondering why we squandered so much investment, and Japan in a positive way, offering to help with infrastructure development.
The currency pegs have a strong impact on forcing investment in the US into sectors that aren't exposed. Until those are resolved, we can either try allocating investment into projects we believe to be good from a central planning standpoint, or we can let the free market allocate the investment. I don't think either is likely to do a good job of that allocation, and don't know which is worse.
Perhaps we impose protectionist measures to halt the export of our inflation. These tariffs may counteract artificially low currency valuations and are plausible (considering you convince people the current Yuan/Yen situation is NOT FREE TRADE).
I think this is plausible too, but the WTO may have a different perspective from ours.
Do you think that the Fed's apparent commitment to inflation, with the help of the treasury, will help to increase inflation expectations and reduce those high real interest rates? I think that this theory may hold water, as Greg Mankiw and Paul Krugman seem to have come to a consensus on the matter (refer to http://krugman.blogs.nytimes.com/2008/12/17/a-whiff-of-inflationary-grapeshot/).
Thanks in advance for your thoughts!
Joe, this is a very sore spot for me. I think this policy has done far more harm than good. I'll write a post on that soon to explain, but in brief, why try to convince everyone that inflation is imminent when that is simply not mathematically possible in the present global configuration? Expectations can affect reality, but reality trumps expectations.
To the extent the inflationary threat is believed, and everyone believes this one, you just create a bunch of scared people with bad information and deal even more damage to the real economy. Uncertainty and fear are a huge barrier for real economic activity.
Since the dollar can't lose value right now by imposition of the pegging countries, and nominal interest rates have hit the zero bound, I conclude that fiscal stimulus and debt monetization will strengthen the dollar and worsen deflation. Until the pegs are broken or serious inflation erupts in the rest of the world, this is the only possible outcome.
Ndk, we have also seen that the US can't handle higher normalized rates (ala 2006 at 5.25%). So, let us imagine we try to raise rates while the economy begins to reflate someday. Eventually, you get to the point where policy becomes restrictive and we end up with an economy flat on its back again, deader than the witch that Dorothy's house landed on.
At some point, a major dollar devaluation will be forced upon us. Hopefully not in the manner FDR autocratically impoverished americans only store of wealth in 1933-1934.
Post a Comment